Key facts
- Google said the TPU 8tx uses hardware-level confidential computing to isolate customer data and model parameters during processing.
- Google claims the TPU 8tx can cut latency by up to half versus TPU 8t with almost no performance penalty.
- The TPU 8i Max combines two TPU 8i chips in one package, doubling computing power, on-chip SRAM and high-bandwidth memory throughput.
- Google is open-sourcing TPU Sync software and claims it can cut certain data-transfer delays and response startup times by more than 90%.
- Amin Vahdat, Google's chief technologist, wrote that TPU 8i Max is built for the memory wall.
- Google's performance figures compare its new systems with earlier Google hardware, not Nvidia's latest products.
Google unveiled two artificial intelligence computing systems on November 9, 2026, the TPU 8tx and TPU 8i Max, adding them to its eighth-generation Tensor Processing Unit lineup. The company said the additions give businesses four configurations for training and running AI models. Google says it serves nearly 100 TPU customers, from AI labs to universities.
Unlike Nvidia's graphics processing units, TPUs are purpose-built AI accelerators. Google developed them for its own services and increasingly sells access to outside companies, including on-premises customers.
Amin Vahdat, Google's chief technologist and senior vice president of AI and infrastructure, wrote in the announcement that computing is moving from human to agent scale and that businesses are asking for accelerators that can reason across large context windows.
Google said the TPU 8tx uses hardware-level confidential computing to isolate customer data and proprietary model parameters even during operation. The system combines TPU 8t silicon with fifth-generation AMD EPYC processors and Google's Boardfly network. Google claims it can cut latency by up to half versus TPU 8t with almost no performance penalty for the added protection. The company said confidential computing cannot prevent a poorly designed AI agent from mishandling data it is allowed to access.
The TPU 8i Max combines two TPU 8i chips into one package, doubling computing power, on-chip SRAM and high-bandwidth memory throughput versus TPU 8i, Google said. Vahdat wrote that the system is built for what he called the memory wall.
Google is open-sourcing TPU Sync, software that moves AI working memory between chips, servers and storage. It claims the system can reduce certain data-transfer delays and response startup times by more than 90%. Its open-source TorchTPU project lets developers run PyTorch and Hugging Face models on TPUs without rewriting applications.
Google's performance figures compare its new systems with earlier Google hardware, not Nvidia's latest products, and the company did not disclose pricing or availability dates.
The announcement came as Nvidia shares fell 9.3% on Thursday, November 5, 2026, and another 5% in Monday, November 9, 2026 trading, after CEO Jensen Huang admitted to an extramarital affair with a woman suspected by U.S. intelligence of spying for China. Nvidia's board has launched an independent investigation, and the company says it has no evidence that corporate secrets were compromised. OpenAI is reportedly planning to purchase Google's TPUs to reduce its reliance on Nvidia, a report the body describes as unconfirmed.